Day: Status Icons and GNOME
Day: Status Icons and GNOME
Posted Sep 1, 2017 19:11 UTC (Fri) by Cyberax (✭ supporter ✭, #52523)Parent article: Day: Status Icons and GNOME
I started reading fully expecting that if a GNOME blog post starts with something like: "We know that this isn’t a good solution. The tray gets in the way and it generally feels quite awkward", then it will end in removing this "something" by the end of the post. I was not surprised at all this time.
Posted Sep 1, 2017 21:04 UTC (Fri)
by johncktx (guest, #113610)
[Link] (18 responses)
Gnome can't be taken seriously at all these days. Hopefully Gnome 3 stops its charade soon and, as gracefully as they can muster, just wind the failed project down and use it as a teaching method for how not to develop a desktop and/or respond to criticism. I once thought it might be able to be salvaged, but those days are long since gone.
Posted Sep 2, 2017 22:27 UTC (Sat)
by sramkrishna (subscriber, #72628)
[Link] (16 responses)
GNOME doesn't believe in sacred cows, and yes, the work we do can be disruptive culturally. It's why I stay on this project for nearly 20 years and enjoy being part of the GNOME community as there is always fresh ideas to work on.
Posted Sep 3, 2017 0:00 UTC (Sun)
by dirtyepic (guest, #30178)
[Link] (8 responses)
Posted Sep 3, 2017 1:38 UTC (Sun)
by pizza (subscriber, #46)
[Link] (2 responses)
Posted Sep 4, 2017 2:45 UTC (Mon)
by bronson (subscriber, #4806)
[Link]
Posted Sep 8, 2017 1:40 UTC (Fri)
by efitton (guest, #93063)
[Link]
Posted Sep 6, 2017 12:51 UTC (Wed)
by jani (subscriber, #74547)
[Link] (4 responses)
Since when has it been okay to tell people what to do, unless you're paying them to do what you want?
The GNOME folks seem to have strong opinions about where they want to take the project, and aren't afraid to make it happen, even if the changes are disruptive. The world is full of projects like that, open source or not. Arguably you need people and projects with strong vision to improve the way we work. Indeed some of the more interesting and disruptive projects are lead by rather opinionated people. (Intentionally not naming any.) Some of the disruptive changes are going to make some users unhappy.
But the people doing the work are free to do so.
You are free to move on to something else. You are free to start or fork or contribute to a desktop environment project with a kernel-like no regressions policy.
Of course, there's also the option of complaining about the state of GNOME 3 on LWN, but isn't that subject growing a tad stale? GNOME 3 is something like six years old now, and there are no signs of it becoming irrelevant despite what you might think based on the comments on this article.
Posted Sep 7, 2017 16:58 UTC (Thu)
by flussence (guest, #85566)
[Link]
Let's rephrase that in gnomespeak[1], then:
“It's time to decide whether GNOME wants to be a useful desktop, a Linux desktop, or a GNOME desktop.”
Posted Sep 8, 2017 1:43 UTC (Fri)
by efitton (guest, #93063)
[Link] (2 responses)
Posted Sep 8, 2017 11:55 UTC (Fri)
by ovitters (guest, #27950)
[Link] (1 responses)
Posted Sep 8, 2017 20:09 UTC (Fri)
by efitton (guest, #93063)
[Link]
What I meant to say:
Obviously different people may have different opinions than me as to why GNOME is frequently a default and different people may have different opinions on the appropriateness of having an experimental and mainstream desktop.
Posted Sep 5, 2017 18:53 UTC (Tue)
by bandrami (guest, #94229)
[Link] (5 responses)
It's default in Fedora, RedHat, and their derivatives. Is it the default anywhere else? It's not default for Mint, Ubuntu, OpenSUSE, Arch, Slackware, Antergos, Manjaro, or Gentoo (to use distrowatch's hit counter as a proxy for userbase). It's one of four defaults for Debian. MATE, Unity, and Cinnamon (all of which are Gnome forks made by people who were unhappy with the direction Gnome has been going) seem to be doing a lot better than Gnome.
A decade ago, Gnome 2 really *was* the default in most distros, and Gnome 3 seems to have thrown that away.
Posted Sep 6, 2017 6:31 UTC (Wed)
by jubal (subscriber, #67202)
[Link] (2 responses)
Posted Sep 6, 2017 9:48 UTC (Wed)
by pizza (subscriber, #46)
[Link]
Posted Sep 6, 2017 13:04 UTC (Wed)
by bandrami (guest, #94229)
[Link]
Posted Sep 6, 2017 9:54 UTC (Wed)
by pizza (subscriber, #46)
[Link] (1 responses)
That made me snicker. Thanks.
> MATE, Unity, and Cinnamon (all of which are Gnome forks made by people who were unhappy with the direction Gnome has been going)
Of those, only MATE is a legitimate "fork".
Cinnamon is all Gnome3 under the hood, only using a different shell to provide a G2-like UI.
Unity was an originally an incompatible fork of prerelease-G3, because (officially) Canonical wasn't willing to wait for G3 to be finished, and because they had specific UI requirements in mind -- not because G3's direction was inherently wrong. Even today, it's far closer to G3 than Cinnamon is.
Posted Sep 7, 2017 3:07 UTC (Thu)
by bandrami (guest, #94229)
[Link]
I definitely enjoyed it, too.
That said: a decade ago, choosing "desktop system" in the Debian installer got you Gnome. Now it gets you a choice of Gnome, the old version of Gnome by another name, the new version of Gnome hacked up to look more like the old version of Gnome, or a desktop that is basically the old version of Gnome but with a much greater willingness to use external components.
Ubuntu coming back into the fold will be a big plus for Gnome, but I still am amazed at how much installbase the team has been willing to give up here.
Posted Sep 6, 2017 18:43 UTC (Wed)
by bronson (subscriber, #4806)
[Link]
Posted Sep 3, 2017 16:30 UTC (Sun)
by jubal (subscriber, #67202)
[Link]
Posted Sep 1, 2017 21:31 UTC (Fri)
by bronson (subscriber, #4806)
[Link] (2 responses)
Posted Sep 2, 2017 20:58 UTC (Sat)
by sramkrishna (subscriber, #72628)
[Link] (1 responses)
Posted Sep 2, 2017 23:57 UTC (Sat)
by bronson (subscriber, #4806)
[Link]
> Another key design principle for GNOME is to put the user in control.
I LOL'd. This article's whole apologetic tone is precisely because it's giving users the shaft. If the GNOME project wanted to put the user in control, it would would first implement (and test!) a solution so this long article wouldn't have to exist. It would just be an entry in the release notes.
> In the next release, we will be introducing a new integration API for file synchronisation apps
Except... the reason you need the API is in *this* release, isn't it? (or am I misunderstanding "next"? Think there'll be unforeseen delays?)
> In GNOME we have two closely related goals: to provide application developers with a clear vision of how apps should be built [that's often changing] and to provide users with a simple, easy to understand and logical experience [with lots of caveats and workarounds].
This section's goal of clear APIs is laudable, but its opening sentence just hangs out there like newspeak. It's at odds with its section and, really, the article as a whole.
(To be clear: I like change. Well-managed change is wonderful. Alas, "Yank it and write some wiki pages for 3rd party devs" is not well-managed change.)
> Many applications today use status icons as a notifications system, despite the existence of the official notifications API, for example.
This part is implying that a number of 3rd party developers are lazy or stupid. Dunno about that... Are you *sure* you understand why they've been reluctant to use the official notifications API?
> We also feel that the consequences of the change won’t be as dramatic as they would have been in the past.
Haha, the GNOME project has been pretty bad at judging how dramatic changes would be in the past, hasn't it? I wonder if things have improved any... (I'll take the under on this one. I predict that GNOME devs will be surprised at how many people are actually affected by this, and how difficult the workarounds described in this article will be in real life. But I do hope I'm wrong!)
> we have actually been using status icons as a crutch for far too long -- that they have been used to fill gaps in our APIs, gaps which are now thankfully getting filled...
Can you picture Allan Day as a physical therapist? He says to his patient on crutches: "You know what? We've been using this crutch far too long." YANK. CRASH. "Don't worry, you'll adjust soon. Most of your leg will heal in a year or two and the rest was obsolete so you won't need it anyway."
Maybe stabilize a solution first, _then_ delete the problem?
But no, alas. This article seems to say that the gameplan is to delete now, think later, and write lots of English to explain about how any pain is actually for the user's own good.
Which, when read in the right tone of voice, is very amusing!
Posted Sep 2, 2017 17:35 UTC (Sat)
by flussence (guest, #85566)
[Link] (24 responses)
Posted Sep 5, 2017 9:09 UTC (Tue)
by aigarius (subscriber, #7329)
[Link] (23 responses)
Posted Sep 5, 2017 9:13 UTC (Tue)
by Cyberax (✭ supporter ✭, #52523)
[Link] (21 responses)
Heck, even Android now has an API to do persistent indicators.
Yet GNOME somehow got it totally wrong.
Posted Sep 5, 2017 14:22 UTC (Tue)
by jubal (subscriber, #67202)
[Link] (20 responses)
Add to this the fact, that many app developers consider status icons a way to enforce – and spread – the brand message, and that the icon images rarely follows design guidelines, and the thing becomes infuriating. (To wit: without bartender my work macbook has SIXTEEN status icons taking roughly one third of the top bar; of those I care only about three; four perhaps.)
Posted Sep 5, 2017 14:23 UTC (Tue)
by jubal (subscriber, #67202)
[Link]
Posted Sep 5, 2017 18:52 UTC (Tue)
by Cyberax (✭ supporter ✭, #52523)
[Link] (18 responses)
I don't _want_ my icon tray to "scale". I'm fine with solution that allows me to have 2-5 indicators for applications chosen by me.
Posted Sep 6, 2017 9:46 UTC (Wed)
by pizza (subscriber, #46)
[Link] (17 responses)
And that's why virtual desktops were commonplace on X Windowing systems at least twenty years ago.
Gnome3 actually improves this paradigm with a dynamic stack-based approach.
> I don't _want_ my icon tray to "scale". I'm fine with solution that allows me to have 2-5 indicators for applications chosen by me.
That, by definition, is "scaling" compared to the overwhelming majority of desktops out there. (Note tha Linux is only a tiny fraction of those)
Posted Sep 6, 2017 11:55 UTC (Wed)
by ken (subscriber, #625)
[Link] (14 responses)
> And that's why virtual desktops were commonplace on X Windowing systems at least twenty years ago.
the dynamic approach is not some sort of universal improvement. I have no idea what problem this solves and it makes using it a pain as you no longer can place windows in the workspace you want as that one may not exist yet.
and the decision to then only have worspaces on the primary display ??? unbelievable! I thought it was a bug until someone pointed out its was done on purpose.
Posted Sep 6, 2017 12:15 UTC (Wed)
by pabs (subscriber, #43278)
[Link] (6 responses)
There is an extension for solving the primary display issue.
Posted Sep 6, 2017 20:51 UTC (Wed)
by mathstuf (subscriber, #69389)
[Link] (5 responses)
By the way, this works because my per-project workspaces tend to be realized through tmux sessions rather than X (or Wayland).
Posted Sep 7, 2017 6:44 UTC (Thu)
by karkhaz (subscriber, #99844)
[Link] (4 responses)
I have six monitors at work, and am using two of them for "per-project" workspaces that are chained together (i.e. when I press <Super-1>, one monitor jumps to workspace 1 and another jumps to workspace 11; <Super-2> changes to workspaces 2 and 12; etc). The other monitors either have a single workspace assigned to them, or (for my web browser monitor) I create and destroy workspaces dynamically. The workspaces on my browser monitor don't have a keyboard shortcut, since there are typically dozens of browser windows open that monitor (each on their own workspace), so I have a program that finds the titles of all of my browser windows, displays them in a dmenu, and whisks me to the workspace holding that browser window. Pure productivity bliss, not counting the millions of hours getting my setup to be this awesome. (Using the i3 window manager, but I'm sure any other tiling WM would work).
But the reason this works so well is that i3 has barely changed its default behaviour since the project started almost a decade ago. The project is mostly adding new features (like workspace saving and configuration options) and fixing bugs.
Posted Sep 7, 2017 7:05 UTC (Thu)
by jem (subscriber, #24231)
[Link] (3 responses)
Tiling window managers are not for everyone. If you value portability in a laptop, then you'll have to compromise on screen size. With a small screen you end up switching between full screen windows.
Posted Sep 7, 2017 9:48 UTC (Thu)
by jubal (subscriber, #67202)
[Link] (1 responses)
Posted Sep 7, 2017 9:59 UTC (Thu)
by karkhaz (subscriber, #99844)
[Link]
It's convenient for me because I can use almost all the same keybindings as on my workstation, although I do see that this is a less compelling argument for folks who only use a laptop---which seems to be more and more people nowadays.
Posted Sep 7, 2017 12:48 UTC (Thu)
by mathstuf (subscriber, #69389)
[Link]
Posted Sep 6, 2017 12:37 UTC (Wed)
by pizza (subscriber, #46)
[Link]
The dynamic approach maps to the way I've always worked -- one desktop per active project. I'm immeasurably more productive with it as it maps to my mental model better than static workspaces. This isn't just my wearing rose-colored glasses either; I'm forced to use a G2-based system at $dayjob, and it's like night and day vs my G3-based personal laptop.
So while I'm not going to claim that the G3 approach is necessarily better for everyone -- yet, for many folks, the G3 approach is vastly superior. (And for those who don't like it, it can be disabled in favor of a static set)
Posted Sep 6, 2017 13:33 UTC (Wed)
by madscientist (subscriber, #16861)
[Link] (5 responses)
On the contrary, that's one of the best features . I put a browser on the secondary (fixed) display and use the primary display workspaces for different types of work. These days you _always_ need a browser available and it's an incredible productivity-killer to have to jump back and forth between workspaces to use it. I can't work well without it anymore.
And of course, if you really don't want it you can disable it as has been pointed out: same with dynamic workspaces (I personally DO disable that and set a static number of workspaces).
Posted Sep 6, 2017 15:46 UTC (Wed)
by ken (subscriber, #625)
[Link] (4 responses)
no its not. its simply wrong. the correct way would be to have all display be part of the workspace and if you wanted to lock some application to be always visible on one display you would simply have an option in the window menu to set it to always display.
then you can have your way of working and everyone else has a sane default and most importantly no need to go in and change some global state for anybody.
Posted Sep 6, 2017 20:48 UTC (Wed)
by mathstuf (subscriber, #69389)
[Link]
Posted Sep 6, 2017 21:07 UTC (Wed)
by corbet (editor, #1)
[Link]
Posted Sep 8, 2017 2:36 UTC (Fri)
by madscientist (subscriber, #16861)
[Link] (1 responses)
No. That's clearly the wrong way to do it. The way GNOME 3 does it is definitely superior.
One of the reasons people don't use multiple workspaces is that it's annoying to have to switch around between workspaces to find things, for cut and paste etc. Having a screen locked means that if you want things to stay always visible you just move things to that screen. This is trivially easy to use and easy to understand for even the least experienced desktop user. You don't even need documentation: it's obvious how it works immediately.
GNOME _does_ have an option in the window menu to set the window to always display, but asking people to figure out how to do it then making them do it every time they open the window is too complicated and annoying. If you learn enough to figure out how to pin a window to the screen, then you're certainly capable of figuring out how to disable the locked screen feature if you don't want it.
Posted Sep 15, 2017 5:24 UTC (Fri)
by flussence (guest, #85566)
[Link]
Posted Sep 6, 2017 17:34 UTC (Wed)
by Cyberax (✭ supporter ✭, #52523)
[Link] (1 responses)
> That, by definition, is "scaling" compared to the overwhelming majority of desktops out there. (Note tha Linux is only a tiny fraction of those)
Every, literally, every other desktop has some kind of tray/menubar indicators: Mac OS X, Android, Windows, iOS (though it's restricted there). Yet GNOME in its great wisdom decided that users don't need them. Facepalm.
Posted Sep 6, 2017 19:36 UTC (Wed)
by jubal (subscriber, #67202)
[Link]
Posted Sep 7, 2017 16:40 UTC (Thu)
by flussence (guest, #85566)
[Link]
Posted Sep 2, 2017 20:55 UTC (Sat)
by sramkrishna (subscriber, #72628)
[Link]
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
1) At least some core GNOME members see GNOME as a place for experimental design. At least some core GNOME members see GNOME as culturally disruptive. This seems pretty well documented, although perhaps not the view of all core GNOME members.
2) At least some core GNOME members desire to see GNOME as a default or the default Desktop Environment. This also seems well documented.
3) My personal opinion (which might be a small minority opinion for all I know) is that trying for both an experimental desktop and being _the_ mainstream desktop is at best inconsiderate to users.
Day: Status Icons and GNOME
You do realise that Ubuntu just dropped Unity, do you?
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Surely. We will all switch to the Unity 8, no?
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
*icon image rarely follows or *icon images rarely follow; grr.
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
> Gnome3 actually improves this paradigm with a dynamic stack-based approach.
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Why choose one or another when you can have both in gnome-shell? :-)
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Day: Status Icons and GNOME
The nice thing is that, if you want that behavior in GNOME, it's a simple configuration tweak away. I agree that having only one display participate in workspaces is weird, but I am happy to flip a switch and get something more usable for me. Everybody should of course set their defaults in a way that pleases me, but I've long since given up on convincing the world of that.
Workspaces
Day: Status Icons and GNOME
Day: Status Icons and GNOME
E17 got this (and so many other things) right, the right way: each screen has independent workspace layouts and switching.
Day: Status Icons and GNOME
And that's why Windows added ability to hide tray icons. Duh.
Windows supports tray management SINCE FREAKING WINDOWS 2000!!
Did I already mention the Bartender app? I think I did. That's how you solve the status icon problem on Mac.
Day: Status Icons and GNOME
Day: Status Icons and GNOME
Millions of people chose GNOME 3 because a standard feature of other platforms, that many apps expect to be there, is badly implemented and now due to be removed entirely? Somehow I doubt that.
Day: Status Icons and GNOME